Building Practical
Machine Learning Systems
B.Tech Data Science student skilled in Python, machine learning, computer vision and NLP, building end-to-end ML pipelines for real-world classification problems.
Find Me Online
Data Science With a Practical Focus
I am a B.Tech Data Science student (lateral entry) with a Computer Science diploma. I build end-to-end machine learning pipelines in Python, from feature engineering and model evaluation to computer vision and NLP models.
I am looking for Data Science and ML roles where I can apply this hands-on experience to real business and classification problems.
Experience & Academic Path
Data Science Job Simulation
Built a Random Forest classifier on 50,000 customer booking records to predict booking behaviour. Used a leakage-safe pipeline with target encoding and 5-fold stratified cross-validation; reached 0.79 test ROC-AUC and 68.2% recall after lowering the threshold to 0.20.
Data Science Internship
Completed a 4-week Data Science internship using Python, covering OOP, NumPy, Pandas and machine learning models.
B.Tech in Data Science (Lateral Entry)
Building on my diploma with a focus on Data Science and machine learning, with hands-on projects in computer vision and NLP.
Diploma in Computer Science & Engineering
Completed a 3-year diploma in Computer Science and Engineering, and started ML work with the traffic sign detection major project.
Interactive Skills & Stack Matrix
NLP & Transformers
Text Models & Fine-Tuning
Computer Vision & Core ML
Visual Perception & Classical ML
Data Analysis & Evaluation
EDA, Statistics & Metrics
Languages & Web Basics
Software Engineering Stack
Featured ML Projects
MCA eConsultation Sentiment Analysis
DistilBERT Sentiment Classifier
Prototype NLP model that classifies eConsultation comments as positive, negative or neutral. Fine-tunes DistilBERT with the Hugging Face Trainer and uses nlpaug text augmentation (synonym, insertion, swap, deletion) for a more robust model.
Traffic Sign Detection in Harsh Conditions
YOLOv8 Object Detection
Fine-tuned YOLOv8s to detect traffic signs. Adding fog, rain, blur, low-light and occlusion images (Albumentations) raised mAP50 from 0.665 to 0.694; a larger 61-class dataset later reached 0.892.
BA Customer Booking Prediction
Random Forest Booking Classifier
Random Forest model that predicts customer booking behaviour from 50,000 records with 5 engineered features and target encoding. Reached 0.79 test ROC-AUC; a 0.20 threshold lifted recall from 16% to 68.2%.
Verified Certifications
Probability, Statistics & NLP
Infosys Springboard
Big Data - 201
Infosys Springboard
Data Science Certificate
NIELIT
BA Data Science Simulation
Forage
Let's Work Together
Open to Data Science and ML roles, internships and collaborations. Send a message below, or email anjha8409@gmail.com.